Bulimia nervosa is an insidious and relatively common disorder, with profound effects on quality of life and mental and physical health. There are some indications that bulimia nervosa may be a persistent disorder with relatively long-term negative consequences. The Specific Aims of this NIMH B/Start application are to determine the ling-term negative consequences. The specific Aims of this NIMH B/Start application are to determine the long-term stability (over 10 years) of bulimic symptomatology. In addition, the predictive utility (over 10 years) of bulimia-related variables--such as perfectionism and drive for thinness--will be determined. Finally, unresolved questions regarding the first-and second-order factor structure of bulimia-related variables will be addressed. The project will analyze data from the 1982-1992 Radcliffe/Harvard Diet and Behavior Project, as well as data currently being collected at Dartmouth college. Overall, approximately 2,500 women have completed assessment instruments on bulimic symptoms and related variables. Approximately 600 of these women participated in both the 1982 and 1992 assessments. The present project will sue multiple regression/correlation techniques to determine the long-term stability and predictive utility of bulimia-related variables ,as well as a host of factor-structure bulimotypic indicators. The project~s findings will have implications for the nature, course, treatment, and prevention of bulimia nervosa